摘要
广泛了解当前智能监控技术已有的运动目标检测算法的基础上，对常用的运动目标几种检测算法进行了深入的分析和对比。针对相邻帧的帧间差分、背景减除法和光流场三种检测方法的不同算法的优缺点，介绍了一种基于背景差分和帧间差分相融合的多帧差分运动目标检测算法。通过对运动目标静态场景监控视频图像的准确获取，运用分析改进算法得到更为接近运动目标实际状况的影像轮廓。
Broad understanding of the basis of the current intelligent monitoring technology has been moving target detection al-gorithm based on, Several commonly used moving target detection algorithm in-depth analysis and comparison. For inter-frame difference of adjacent and frames Background subtraction method and optical flow, Advantages and disadvantages of different al-gorithms methods, Introduces the multi-frame difference moving target detection algorithm based on background subtraction and integration of inter-frame difference.By moving target surveillance video images of a static scene accurate access, Analysis us-ing improved algorithms get closer to the actual situation of the moving target image contours.
作者
李立现
LI Li-xian （Information Science and Technology Institute of Jiujiang University, Jiujiang 332005, China）
出处
《电脑知识与技术》
2013年第12期7864-7866,共3页
Computer Knowledge and Technology
基金
九江学院校级科研课题（课题编号：2011KJ11）
关键词
运动目标检测
光流法
背景差分法
帧差法
Moving Target Detection
Optical Flow
Background Subtraction
Frame Difference Method
